An ops layer vs a managed hosting service
ClickHouse Cloud is a genuinely zero-ops way to run ClickHouse — autoscaling, backups, upgrades handled for you, plus its own Ask AI agent and a remote MCP server in beta. chmonitor isn't a hosting product; it's an independent monitoring and advisor layer that works on any ClickHouse — including ClickHouse Cloud itself. They're not mutually exclusive.
chmonitor vs ClickHouse Cloud's built-in tooling
ClickHouse Cloud legitimately wins on zero-ops managed hosting — that's not something chmonitor does either.
| chmonitor | ClickHouse Cloud | |
|---|---|---|
| What it is | An independent monitoring/ops layer you run alongside any ClickHouse deployment | ClickHouse Inc's fully managed ClickHouse hosting service |
| Fully managed hosting (no servers to run) | Not a hosting product — it's an ops layer on top of whatever ClickHouse you run | Zero-ops managed service: autoscaling, backups and upgrades handled for you |
| Works with self-managed / on-prem / any-cloud ClickHouse | Any ClickHouse — OSS, Altinity, ClickHouse Cloud — on your infra or any cloud | Monitors the ClickHouse Cloud service itself, not clusters you self-manage |
| ClickHouse-specific ops depth | Dedicated pages for the full merge/part/replication-queue/Keeper lifecycle | The Cloud console shows service-level metrics and Query Insights, with less part/merge/replication drill-down than a dedicated ops tool |
| AI agent / natural-language assistant | Built-in agent + MCP server, works against any connected ClickHouse — analytics and cluster ops | Ask AI agent (public beta, March 2026) — natural-language analysis over your Cloud service's data; analytics-first, not a cluster-ops advisor |
| MCP server for AI tools | Ships with every deployment, self-hosted included | Remote MCP server in public beta (March 2026), Cloud-only, OAuth-gated, read-only SELECT scope |
| Query advisor (recommend-only) | Ranked index/rewrite suggestions with EXPLAIN-based impact estimates; never auto-applies DDL | Ask AI can write and run queries, but isn't a schema-optimization advisor |
| Materialized view / projection designer | Recommend-only MV & projection suggestions derived from real query patterns | No dedicated MV/projection design tool |
| Anomaly detection on cluster health | Statistical baselines on cluster/query metrics, built in | Console surfaces usage and service metrics; no dedicated statistical-baseline alerting product |
| Alerting & webhooks | Cluster-health thresholds and outbound webhooks run everywhere; alert history and inbound-event storage persist with a D1 binding (hosted cloud or a Cloudflare Workers self-host) | Billing/usage and service notifications; not general cluster-health threshold alerting with webhook delivery |
| Pricing model | Free self-hosted (GPL-3.0); hosted cloud in early access | Usage-based: ~$0.22–$0.39 per compute-unit-hour plus ~$25/TB-month storage; no permanent free tier (30-day / $300-credit trial) |
| Open source | GPL-3.0, full source available | Proprietary managed service (built on the open-source ClickHouse core) |
Pick your hosting, keep the ops advisor
chmonitor doesn't compete with ClickHouse Cloud's hosting — it adds the ops-focused advisor and alerting layer on top, wherever your ClickHouse runs.
Not locked to one hosting vendor
Works against self-managed ClickHouse, Altinity, or ClickHouse Cloud — connect any of them, switch hosting without losing your monitoring.
Built for cluster ops, not just queries
A recommend-only query/MV advisor and statistical anomaly baselines built to run a cluster, not just analyze data in it.
MCP server, every deployment
Self-hosted chmonitor ships the same MCP server ClickHouse Cloud offers only in beta, and only to its own managed service.
Free self-hosted, period
GPL-3.0, no usage-based bill. ClickHouse Cloud bills compute + storage with no permanent free tier.
More comparisons: chmonitor vs Grafana · chmonitor vs Datadog · ClickHouse vs TimescaleDB · ClickHouse vs Postgres · ClickHouse vs Druid vs Pinot
Start monitoring — advisor and alerts built in
Open the hosted dashboard and connect a cluster, or self-host with Docker, Kubernetes, or from source. Self-hosting is always free.